Yves here. This factory building trend, sighted by Wolf Richter, bears watching. It’s a concrete demonstration (pun intended) that there is a serious effort underway to “reshore” manufacturing. However, desire and spending do not necessarily translate into results. Recall the great hyped Foxconn factory in Wisconsin, where they sucked lots of subsidies out of the state and then punted. From CNBC in 2021:
Taiwan electronics manufacturer Foxconn is drastically scaling back a planned $10 billion factory in Wisconsin.
Under a deal, Foxconn will reduce its planned investment to $672 million from $10 billion, and slash the number of new jobs to 1,454 from 13,000.
And remember, this is an established player, as in Foxconn knows how to start up and operate factories. This is not Americans who might be scrambling to find desperately needed factory floor supervisors and higher level manufacturing managers.
Wisconsin did at least restructure the deal and clawed back a lot of the subsidy commitments.
Similarly, Alexander Mercouris has reported long form on the failed US effort to increase 155 mm shell production. In addition to many mishaps which he recounted long form, the bottleneck remains a lack of gunpowder. There seems to be only one factory, in Poland, that makes TNT. The US had decided to move off TNT because environmentally nasty, but the replacements never worked (not sure because they failed as explosives or were too hard to produce at scale and/or reasonable cost).
Admittedly, it’s a somewhat different type of manufacturing, but I was a paper mill brat and my father ran paper mills. Coated paper is very fussy. The coating, in his day mineral clay, is applied when the paper is wet. The paper machines have to operate at very fine tolerances or else the paper breaks, causing costly downtime.
The mills have to run 24/7 except for scheduled maintenance because the capital costs are high.
My father ran startups and turnarounds because hardly anyone in the industry could execute them. He eventually became the head of manufacturing at one of the major papermakers.
The rule of thumb was a successful startup (of a single “machine” as in production line, which cost ~$500 million in 1970s) took two years and cost 20% of the capital costs. An unsuccessful startup was a running money sore.
In other words, erecting factory buildings is the easy part. Stay tuned as to how high the successful opening and production rate is.
By Wolf Richter, editor at Wolf Street. Originally published at Wolf Street
Companies invested a record $19.7 billion in June in the construction of manufacturing facilities, up by 18.6% from the already surging levels in June 2023, up by nearly 100% from June 2022, and up by 209% from June 2019, according to the Census Bureau today.
The investment totals here only cover the actual construction costs of the facilities, not the costs of the manufacturing equipment and installation that can dwarf the construction costs of the building. The total cost of a big chip plant might reach $20 billion, but the construction costs are the smallest part of it. So the total amounts invested in manufacturing plants, including the equipment and installation, are much higher. But here, the amounts only refer to the construction of the plants, and can be seen as a directional indicator of total investment in manufacturing.
In addition to the construction boom of semiconductor plants, a large number of other manufacturing plants have been announced, and continue to be announced.
The explosion in factory construction that started in the second half of 2021 was one of the changes that came out of the pandemic when America’s scary dependence on China became apparent in massive shortages of all kinds of goods, including semiconductor shortages, and unbelievable supply-chain and transportation chaos, that caused corporate America and policy makers to rethink the strategy of endless globalization.
The CHIPS Act, signed into law in August 2022, was part of the movement. While the first awards have been announced, there is lots of stuff left to do, including due diligence, and the cash hasn’t been disbursed yet. That’s still to come.
The 12-month total of investment in manufacturing plants jumped to $235.5 billion, up by 19% from the same period a year ago, up by 100% from two years ago, and up by 217% from the same period in 2019.
Construction spending on manufacturing facilities now accounts for over 10% of total construction spending in the US, residential and non-residential, from single-family houses to roads and power plants.
It’s all based on the principle that industrial robots cost the same in the US and China, that manual labor is a much smaller cost component in modern automated manufacturing, and that transportation costs (which spiked during the pandemic) and loss of Intellectual Property (IP), which is a given in China, and other risks have to be added to cost equation.
In addition, the increasingly complicated and stressed relationship between the US and China has exposed for all to see that the reckless dependence by US companies on production in China is a fundamental risk, not only for the companies, but also for national security.
No one is going to build a factory in the US to make low-value products, such as T-shirts. It’s all focused on complicated high-value products, such as motor vehicles, chips, electrical and electronic products, heavy components and equipment, etc.
Inflation in Construction Has Abated
The Producer Price Index for construction costs of nonresidential buildings, after blowing out in mid-2021 through 2022, started plateauing in early 2023 and has remained roughly unchanged since then (red in the chart below).
On a year-over-year basis, the PPI for nonresidential construction has been flat to slightly negative since late 2023, after having spiked by as much as 24% in mid-2022 (green).
I am certainly not an expert, but it seems to me this is probably driven by dumb institutional money looking to make a formula allocation to finance commercial real estate development. Office and multifamily are now way overbuilt, and mall-scale retail is hardly a thing anymore, so industrial is the least toxic looking option. We’ll see in another year or two whether the producer companies actually start filling these shells with expensive equipment and skilled operators.
To a certain degree I think this is part of a worldwide phenomenon, where increasing concerns about globilization has led to a huge (usually government led) increase in direct industrial investing. Its not just China – countries like Indonesia and Vietnam are investing enormously in direct industrial capacity in both ‘old style’ industries like steel production as well as all trying to hop on board the EV/battery train.
The likely result will be a huge glut in manufacturing capacity in the next major downturn. There will be winners and losers, but a likely outcome will be a major increase in protectionism worldwide. While trade is often seen these days as a China vs US thing, its a lot more complicated – many significant Asian countries are trying to follow the Japan/ROK/China playbook and are not happy about facing what they see as unfair competition. EU/US trade is still enormous and particularly important to Europe. The problem of course is that everyone wants to export, but nobody wants to import. A vast amount of money is at stake and its hard to escape the conclusion that a lot of countries are banking on someone else buying their surplus capacity. Someone will get a nasty surprise.
One point in the US’s favour – while there has undoubtedly been a loss of institutional knowhow in US production, relatively high wages and easy immigration allows the US to import know-how at a scale no other country or region can match. There is a huge worldwide industry involved in the process of getting manufacturing from greenfield to production (lots of established companies are quite bad at this, and they know it, so they hire in expertise for latchkey projects), and that industry is by necessity highly mobile. The engineers and technicians involved in this are well used to hopping on planes to spend a couple of years somewhere different to set up yet another new factory/refinery/chemical plant or whatever. They will be much in demand over the next few years, and I suspect many of them will find themselves in the US.
Semiconductor factories emit enormous amounts of CO2, deplete water sources for farmers, cause health problems for factory workers, and pollute the waters around the factories with heavy metals. Is America ready to accept that?
https://www.theguardian.com/environment/2021/sep/18/semiconductor-silicon-chips-carbon-footprint-climate
As long as printing money remains our biggest industry, little real production would seem worthwhile to do. Our expected trade deficits are now $4 trillion/year. That represents free stuff right there, not just through cheap foreign labor but the materials, the transport, the insurance – all of it – for free. You can’t compete with free even if that means living with a risky supply chain.
As an update on the Foxconn debacle, Microsoft has stepped in and is constructing a datacenter and training facility on part of the land. The effort is led by Microsoft President Brad Smith, who’s originally from Appleton, WI and is imho sincere about making a significant, long term investment in his home state. Microsoft also promised on-the-job training for 100k workers as part of the deal. Microsoft is buying up parcels and has actual blueprints approved for what the final project will look like. I’m no Pollyanna but this looks like a pretty good result for what was otherwise a total boondoggle. There are a series of stories in the Milwaukee Journal-Sentinel describing just what was going on at Foxconn. It seems there was a change in their business plan that was not communicated to Wisconsin officials and the people in Taipei just thought they could wait it out. One theory is that they counted on the statute of limitations on fraud running out, when they could just cut and run.
https://www.jsonline.com/story/money/business/2022/04/07/foxconn-stalling-development-dodge-30-million-annual-tax-bill/7266263001/
US manufacturing productivity has been declining since 2011, or for the last 12 and a half years
https://fred.stlouisfed.org/graph/?g=naRn
January 30, 2018
Manufacturing Production and Manufacturing Productivity, * 1988-2024
* Output per hour of all persons
(Indexed to 1988)
https://fred.stlouisfed.org/graph/?g=m2mx
January 30, 2018
Manufacturing Productivity, * 1988-2024
* Output per hour of all persons
(Percent change)
https://fred.stlouisfed.org/graph/?g=m2mB
January 30, 2018
Manufacturing Productivity, * 1988-2024
* Output per hour of all persons
(Indexed to 1988)
https://fred.stlouisfed.org/graph/?g=naRA
January 30, 2018
Manufacturing Production and Manufacturing Productivity, * 1988-2024
* Output per hour of all persons
(Percent change)
Spending money without a comprehensive strategy seems futile.
Any serious push to onshore manufacturing needs to include education. For example, machinists. Human capital more than physical plants. A really serious strategy would institute single payer insurance and elimination of private health insurance and PBMs, for starters, to reduce labor costs. Being the US, do these as part of the war mobilization for China proposed in the recent defense strategy assessment.
“…However, desire and spending do not necessarily translate into results…”
This is going to be very interesting going forward. State subsidies, “incentives”, tax breaks combined with Federal CHIP grants and subsidies can add up to many billions.
States like Oregon may have to make exceptions to environmental regs. As noted above, the environmental impact of chip factories is very big. I’m not clear on how that would work.
Intel announced layoffs, suspension of dividends, and introduced retirement incentives recently, at the same time they are getting large subsidies and “incentives” https://apnews.com/article/intel-chip-ai-job-cuts-layoffs-loss-e61781e9364b69af63481c34ca5dcd67
I would keep an eye on that as well, to be crude: it might be a great way to swindle billions from the public purse.
Plutonium Kun makes a good point about advantages that the US could enjoy. Plus, in light of increasing tensions with China over the Taiwan issue, the US can politically “rationalize” massive subsidies to make the US more self-sufficient in these areas – national security and al that. Plus it is touted as a huge “job creator”. What’s not to like?
One negative factor is cost of living in the USA. Chip mfg. states like Oregon, have high real-estate and rent costs, and in the USA as a whole, so-called health care is the most expensive in the world, by a significant margin. Very high overhead costs may well have a medium term impact after all the subsidies and construction is over
To be typically skeptical: it remains to be seen if this will all play out as advertized.
So with all this frantic planning based on competition for gain and profits and ultimately aggression worldwide, it would make better sense to create a new forum whereby planning was the means to create cooperation and reduce waste and pollution. And provide for conservation of resources both natural and social. Highways and stoplights instead of high-speed guard rails and vast junk yards.
The article doesn’t go into specifics on what type of factories are being planned. Just vaguely mentions high end manufacturing. Anyone have ideas on specifics? Or it is all chips?